Removing sun glint from optical remote sensing images of shallow rivers

Earth Surface Processes and Landforms
By:  and 

Links

Abstract

Sun glint is the specular reflection of light from the water surface, which often causes unusually bright pixel values that can dominate fluvial remote sensing imagery and obscure the water-leaving radiance signal of interest for mapping bathymetry, bottom type, or water column optical characteristics. Although sun glint is ubiquitous in fluvial remote sensing imagery, river-specific methods for removing sun glint are not yet available. We show that existing sun glint-removal methods developed for multispectral images of marine shallow water environments over-correct shallow portions of fluvial remote sensing imagery resulting in regions of unreliable data along channel margins. We build on existing marine glint-removal methods to develop a river-specific technique that removes sun glint from shallow areas of the channel without overcorrection by accounting for non-negligible water-leaving near-infrared radiance. This new sun glint-removal method can improve the accuracy of spectrally-based depth retrieval in cases where sun glint dominates the at-sensor radiance. For an example image of the gravel-bed Snake River, Wyoming, USA, observed-vs.-predicted R2 values for depth retrieval improved from 0.66 to 0.76 following sun glint removal. The methodology presented here is straightforward to implement and could be incorporated into image processing workflows for multispectral images that include a near-infrared band.

Additional publication details

Publication type Article
Publication Subtype Journal Article
Title Removing sun glint from optical remote sensing images of shallow rivers
Series title Earth Surface Processes and Landforms
DOI 10.1002/esp.4063
Volume 42
Issue 2
Year Published 2017
Language English
Publisher Wiley
Contributing office(s) National Research Program - Central Branch
Description 16 p.
First page 318
Last page 333